Autonomous navigation using an adaptive hierarchy of multiple fuzzy-behaviors

Abstract

Adaptive behavioral capabilities are necessary for robust mobile robot navigation in non-engineered environments. Robust behavior requires that uncertainty be accommodated in the robot control system, especially when autonomy is desired. Fuzzy logic control technology enables development of controllers which can provide the necessary computational intelligence in real-time. This paper presents the incorporation of fuzzy logic, into the framework of behavior-based control. An architecture for hierarchical behavior control is presented in which control decisions result from a consensus of behavioral recommendations. Multiple fuzzy-behavior coordination is discussed and applied to autonomous navigation without explicit maps. Performance and robustness is demonstrated by implementation on a mobile robot with significant mechanical imperfections.